Stone Strike and Shuriken Mastery mysteries revealed!

Arvec · 15 Aug 2006

Let's start with the boring stuff - the stuff that everyone should/does know:

But first, a message from our sponsors:

This research was proudly funded by the SOS Brigade, whether we(they) knew it or not - the money for the excessive amount of pots came out of what I was going to donate

And now back to our research.

Stats that effect damage:

Stone Strike: DA
Shuriken Mastery: DA

Rumors that AP has an effect on Stone Strike are just that, rumors. This was determined by removing my weapon and attacking with stone strike, then equipping a weapon with +80AP and using stone strike against the same enemy again. The damage numbers yielded were the same. Obviously, equipping a weapon with higher DA yielded higher damage numbers. This was never in question.

This is where the mystery begins! A lot of people had their own views and opinions on how the accuracy of these attacks was determined. With grim determination, and a lot of free time (it was my night off), I put myself on the case. Fine, it was a mandatory assignment, and I hated it. Or did I? At any rate.

Stone Strike!
There have been a lot of rumors as to what stat effects the accuracy of this skill. To put it bluntly, rumors suck. So here are some facts, backed with research.

There were four samples taken, each one with 200 hits vs the same target, but isolating certain stats by use of equipment.

L45vL64 - Galli, path to phantom school (This is consistent between all samples)

Base stats:
This test is simply to give us a benchmark to compare the other samples to. All gear that gave a bonus to the stats in question was removed.
DA: 22
LK: 11
AC: 52
Total hits: 124 (62.00%)
Total misses: 61 (30.50%)
Total crits: 15 (7.50%)

Absolute base hit% with listed stats: 69.50%

+8 Luck: (luck accessory, penguin pet)
This test is to see if luck has any effect on hit rate directly, or if it raises hit rate indirectly by raising critical rate (if a critical hit is automatically a hit).
DA: 22
LK: 19
AC: 52
Total hits: 120 (60.00%)
Total misses: 60 (30.00%)
Total crits: 20 (10.00%)

Absolute hit% with +8 Luck adjustment: 70.00%

Conclusions: There was a sizeable increase in critical rate (2.5% absolute, or +33% relative increase over the base stats), however, the absolute number of hits was roughly the same - enough to attribute the difference to randomness. The total number of hits was decreased, while the total number of criticals increased by almost the same amount. This not only suggests that luck does not increase hit rate, but that criticals are not automatically counted as hits, or that the critical chance is only checked after a confirmed hit.

+8 AC: (attack armlet, +2 AC pet, +2 AC sword)
This is to test the effects of AC on stone strike hit chance.
DA: 22
LK: 11
AC: 60 (recall that base AC is 50)
Total hits: 136 (68.00%)
Total misses: 47 (23.50%)
Total crits: 17 (8.50%)

Absolute hit% with +8 AC adjustment: 76.50%

Conclusion: Hit rate was greatly increased over the base stats - a 7% bonus, or about a 10% increase compared to base stats. The critical rate was also increased at roughly the same ratio as the normal hit rate. However, it should be noted that to increase the AC by this amount, it will usually require a significant reduction in DA from bonus points or gear, or alternatively, require a sacrifice of survivability based on build.

Updated note: Please do not take this as meaning 8 AC will yield a 6.5% increase. As pointed out by BrTarolg, the sample size isn't large enough to give close estimates on impact - rather the intention is simply to determine if it has an effect. Since the difference is so large when compared to all of the other samples, I can confidently say that AC has an effect on Stone Strike accuracy.

+15 DA: (+4 DA weapon, sixth sense)
Do I need to specify the reason for this?
DA: 37
LK: 11
AC: 52
Total hits: 125 (62.50%)
Total misses: 59 (29.50%)
Total crits: 16 (8.00%)

Absolute hit% with +15 DA adjustment: 70.50%

Conclusion: Hit rate was again, roughly the same as the base. The amount of criticals did not stand out, either. While DA is the determining factor in damage, it is concluded that it does not have any effect on stone strike accuracy, or that it's impact is extremely small, so much as to be unnoticeable - especially since the bonus to DA was more significant than the bonus to the others (assuming growth from base stats, IE AC is counted as 2, not 52).

Overall conclusion: While luck did increase the critical rate, there was no increase recorded in the hit rate of Stone Strike. The overall hit% with +DA was roughtly the same as that of the base stats, as well, enough to be attributed to randomness. The only stat that had an impact in hit rate large enough that there was no doubt it could be attributed to randomness was AC. To be short...

The accuracy of Stone Strike is determined by AC alone!

And now for something completely different.

Shuriken Master!

Shuriken master was a much quicker test. Why? I'll start by showing you the results of the base stat test, coincidentally, the only test...

L46vs64 (Galli again - poor Sanguine Mantles...)
DA: 22
LK: 11
AC: 52
Hits: 86
Misses: 0 (!)
Crits: 4
Notice that the sample size was only 90. After 90 consecutive attacks with NO MISSES I abandoned this test, with a new respect for shuriken master, and with near-certainty that...

The accuracy of Shuriken Master is determined primarily by DA!

...or that it is based on an amount of luck, or a luck/DA combination so small that the base stats of any sense type is enough to ensure an extremely high hit-rate. Note that this does not mean that the accuracy of shuriken master is 100%! You WILL eventually miss with it. But the high base accuracy of shuriken master allows no room for testing, as the hit rate is simply too high for any stat alternation to have significant effect. Not that this is a bad thing...

And so, the mystery is solved (hopefully). I'm also working on the great mysteries of what lies under the depths of caballa island. And by that, I mean drilling up a storm. Expect that one Soon (TM).

Disclaimer: We (I) reserve the right to abandon any research into the mysteries of "the unknown" on a whim, with flimsy excuses such as "I was bored", or "it was hard".

Hide · 15 Aug 2006

May i know whats is your level? I have 55 DA at level 50, thats with sixth sense on.

Pumpkin · 15 Aug 2006

wow thanks for all the work that went into the test. It has really cleared up some question.
lol *note to self ..... stop adding luck .
Can you please keep us informed if you fight tougher monsters and start missing more with Shuriken master. Thanks

BrTarolg · 15 Aug 2006

hi this is great work and all...

but im really sorry to say... if you consider a 2.5% increase in crits a "significant increase", or to say 8 extra AC gives you an extra "6.5% accuracy" then thats just a wrong assumption to make.
the sample size is WAAY to small to narrow it down to .5%. if you want to narrow it down to a WHOLE 5% <give or take a percent>, youd need to do a sample of at least 1000. saying that you hit 76.5%, could actually mean anywhere from hitting 80% to hitting the "usual" 70%.

however i will say that is is now SAFER to assume from what youve done <sorry to be harsh youve put in great work. i suggest you keep the results because you can add them up together later on> that in general with your build, you hit around 70% with stone strike.

why? because youve done 600 trials of it. this means its MUCH safer to narrow it down that way.

----

anyways, i have also been recording my stats, and recent stone strike hits also. maybe later when im finished, i can post up my results and compare them to yours. <and most likely, add them together and average them out>

unfortunately, i dont have a way of increasing my AC yet. but i will do that later.

Arvec · 15 Aug 2006

Originally Posted by BrTarolg

hi this is great work and all...

but im really sorry to say... if you consider a 2.5% increase in crits a "significant increase",

Sorry, should have been more specific. It was a significant increase relative to the critical percentages achieved with the base stats - a 33% relative increase, if I remember correctly.

Originally Posted by BrTarolg

or to say 8 extra AC gives you an extra "6.5% accuracy" then thats just a wrong assumption to make.

That wasn't exactly the purpose of this. It was merely to determine which stat(s) had an effect on stone strike accuracy, and I believe that was accomplished, despite the moderate sample size. The relative increase to accuracy with the extra AC bonus was roughly 10% from the base - I believe that is significant enough that the result would not be skewed by randomness. While it's possible, it seems..unlikely. I will admit there's a possibility that it's wrong due to randomness simply throwing a curveball. But that's how statistics are. Also, in most cases, you can't really determine so much the exact bonus of accuracy - I'm assuming that the effects per point of accuracy provide a sort of bell curve effect. That would especially be true as accuracy becomes extremely high, as most games simply don't give you a 100% hit chance (though some simply hard-code it in), so having an extremely high AC would have less of an effect per point. This is why I scrutinize guides that give you a percentage increase per point of accuracy. That's not what I was trying to do.

Originally Posted by BrTarolg

the sample size is WAAY to small to narrow it down to .5%. if you want to narrow it down to a WHOLE 5% <give or take a percent>, youd need to do a sample of at least 1000. saying that you hit 76.5%, could actually mean anywhere from hitting 80% to hitting the "usual" 70%.

As I said, I wasn't really aiming to give the exact effects of x points of AC, but rather just determine the trends, or maybe provide ballpark figures or estimates. The .5% is just what the math came out to.

I don't take any offense at what you said, but rather look at it as a way to refine the work that has been done, and clarify the exact intentions.

I added a note below the conclusions on AC specifying that the results aren't to be taken 'as gospel' and that the sample size isn't enough to give clear conclusions as to the exact effect of accuracy. Thanks for pointing this out.

BrTarolg · 15 Aug 2006

well ok ive done another test so you can add this to your results.

i was lvl 31, i had a DA of 14+13, and a AC of 51

these were against lvl 31 <torpedo fish> and lvl 37 <merrow> monsters

i was using six sense, and skill stone strike

HIT:310
MISS:126

edit: woops i madea mistake. no its exactly the same as you have, 70% hit and 30% miss :P

tbh, i reckon that the majority of the hit percentage comes from the level of the character, not the AC

i mean, i had AC 51 which was the same as you right? but i was low lvl, only 31, killing levels of 6 higher than me. the vast majority of the monsters are fighting are merrow.

so what level were you, and what lvl monsters were you fighting?

Arvec · 15 Aug 2006

The level gap was nineteen levels (45v64). I'd try to quantify it, but there was a difference in the AC as well as the level gap being much narrower. It's impossible to quantify with different stats, really. I'd run the accuracy test against something lower level, but there's also the possibility that the enemy I choose will have a higher or lower evade, thus skewing the results.

Wait....I'll do it again when I'm later level. Unfortunately, my AC may be higher by then...but I did remember to save the equipment, in the event I wanted to take more samples. I may actually have to make a +1 AC gear to compensate, in case I level up in that time. At the very least, I seem to have confirmed that the other stats to not effect accuracy, so I can maybe use a higher sample size. But I'll have to determine the entire base benchmark again, to ensure that it hasn't changed as I levelled >.>

As far as combining data, it's not really possible. There are any number of factors. Relative to my samples, the enemy you fought may have had lower or higher HV. Level difference could have an impact. That one point of AC could have factored in. In fact, all of these could have combined to give you the same result, when really, in the same ideal situations, they may have been quite different. In fact..

It appears you underestimated your hit% (or made a typo in attacks or percentage).
Total attacks: 436 (310+126)
Hits: 310
310/436 = 0.7110091743119266055045871559633

70% vs 71% is close enough for random deviations, and a level based modifier may even explain the difference, as the level gap was narrower than my testing, though the base AC was actually slightly higher as well. The existence of a level based modifier may be worth investigating. If I get bored.

Edit: I also just realized you used two different monsters in this test. As mentioned before, they may have completely different sets of 'hidden' stats, or there could be any number of nuances in the formula that we simply don't know. The only way to get a lock on these things is to ensure consistency to the highest degree outside of what you are examining. Of course, since you brought up the issue in the first place, you probably already know this, just didn't realize it at the time. I feel I might have rushed you into it, while I had been planning this for a few days before I even made the excel spreadsheet.

BrTarolg · 15 Aug 2006

when i say combining data i dont literally mean to combine the data

its just more stuff to look at.

the fact tht im still hitting ~71% of the time is quite interesting

however, i still beleive that going pure DA is the best option. if i added 10AC, even if i had an extra 10-15% increase in accuracy, i do not think will compensate for the damage lost from less DA.
assuming that i will hit on average 70% of the time, if i do 1k damage, after 1000 uses of SS, that will be around 70,000 damage
add just 1 DA extra to that, and i will be doing an extra 2800 damage for every DA i add <as far as i remember, 1 DA = 40 extra damage on SS>
lets be generous and say 1 AC adds 1.5% accuracy, so that will only add another 1500 damage, which is quite alot less.

----

actually what i wrote is flawed. it completely depends on how much damage DA actually adds. i remember reading a few times in different places it adds 40 damage but ill have to check.

in fact, ill go ingame and do it right now
-------------------------------------------------
ok results as follows -

lvl 10 SS, 16 DA, vs fanta fish = 758 damage
lvl 10 SS, 16+6 DA vs fanta fish = 1034 damage
lvl 10 SS 16+14 DA vs fanta fish = 1402 damage

in all these cases, each point of DA adds approx 47 damage. the reason why its not exactly 47 is probably due to rounding errors <because SS adds a percentage damage to my attack>

whats interesting tis that i do 276% damage... but of what? is it suggesting that my "base" damage is calculated by 17xDA?

16 DA, vs shell trap = 777 damage
16+6 DA, vs shelltrap = 1053 damage
16+14 DA, vs shelltrap = 1421 damage

ok here again it looks more like each DA adds 47~ or so damage.

either way, if it adds THAT much damage per attack, it probably wouldnt be worth wearing AC unless each point gave us like.. a 5% bonus or something :P

kinoish · 15 Aug 2006

hm.. i recall a lion name furlove wrote somewhere on this forum that accuracy is not the contributing stat for SS's hit rate because he puts all points into AC, and he still misses alot whenever hes using SS. i think he said he used SS to save money or something... ~.~;;

anyway great job arvec & brtarolg, im looking forward to reading more of your tests.

oryx · 15 Aug 2006

acutaly, I'm surprised you haven't come to teh conclusion I have... that ss HAS A FIXED HIT RATE.... and this is how I came to that conclusion.

I miss all the time... on level 7 monsters. I don't miss with anything BUT SS. so therefor, since all my stats at level 60 are FAR beyond what I need to hit a level 7 monster, I conclude that the hit rate must be fixed (apparently around 70% XD)

thank you SOOOOOO much for alll of your hard work gguys.... it took me FOREVER to come to my conclusion, and I wish I'd had something like this to read first XD

Arvec · 15 Aug 2006

Originally Posted by BrTarolg

when i say combining data i dont literally mean to combine the data

its just more stuff to look at.

the fact tht im still hitting ~71% of the time is quite interesting

however, i still beleive that going pure DA is the best option. if i added 10AC, even if i had an extra 10-15% increase in accuracy, i do not think will compensate for the damage lost from less DA.
assuming that i will hit on average 70% of the time, if i do 1k damage, after 1000 uses of SS, that will be around 70,000 damage
add just 1 DA extra to that, and i will be doing an extra 2800 damage for every DA i add <as far as i remember, 1 DA = 40 extra damage on SS>
lets be generous and say 1 AC adds 1.5% accuracy, so that will only add another 1500 damage, which is quite alot less.
<more stuff about DA>
either way, if it adds THAT much damage per attack, it probably wouldnt be worth wearing AC unless each point gave us like.. a 5% bonus or something :P

Yeah, I'm pretty sure I made a note that it's not really worth raising, but I'll look over it again. Or maybe I no longer have to?

I am also of the opinion that the accuracy bonus is not worth it when compared to the loss of damage.

Originally Posted by oryx

acutaly, I'm surprised you haven't come to teh conclusion I have... that ss HAS A FIXED HIT RATE.... and this is how I came to that conclusion.

I miss all the time... on level 7 monsters. I don't miss with anything BUT SS. so therefor, since all my stats at level 60 are FAR beyond what I need to hit a level 7 monster, I conclude that the hit rate must be fixed (apparently around 70% XD)

thank you SOOOOOO much for alll of your hard work gguys.... it took me FOREVER to come to my conclusion, and I wish I'd had something like this to read first XD

If you have a 1XXX build as well, your accuracy is currently 54, correct? So your accuracy would probably only be a bit higher than 70%, even against low level enemies, since we have such a meagre amount of AC to begin with. What you've said indicates that there is not a level-based modifier, but I'll still test it out later.

And hey, if it were hard work, I wouldn't be doing it! Anybody can do it, they usually just need a little motivation. Well, anybody can do it with the exception of those that haven't learned the math in school yet. That would be a bit of a problem

Malicx · 15 Aug 2006

While I was browsing the japanese trickster wiki I noticed that there are formulas given for the accuracy of both attacks (at least, that's how I understood it after google translated it :P)

Stone Strike on-target impact on-target hit ratio =72+MOB evasion ×2.5 (fractions omission)

Assassin Throw on-target impact on-target hit ratio =39.5+MOB evasion ×2.5 (fractions omission)

I then looked around for a "MOB evasion" stat, but never found it. I'm still intrigued by the formulas, though.

Arvec · 15 Aug 2006

Ah? They don't really make sense. Why would it be an addition of the mob evasion to the base hit chance? And why would shuriken mastery have a lower base chance to hit? It doesn't include any stats, either, suggesting a flag hit-rate. However, accuracy has shown to have a higher hit-rate. At any rate, I'll be checking it again at higher level to test the existence of a level-based modifier.

BrTarolg · 16 Aug 2006

Originally Posted by Malicx

While I was browsing the japanese trickster wiki I noticed that there are formulas given for the accuracy of both attacks (at least, that's how I understood it after google translated it :P)

Stone Strike on-target impact on-target hit ratio =72+MOB evasion ×2.5 (fractions omission)

Assassin Throw on-target impact on-target hit ratio =39.5+MOB evasion ×2.5 (fractions omission)

I then looked around for a "MOB evasion" stat, but never found it. I'm still intrigued by the formulas, though.

i think its adding to the mob evasion to calculate how often we will miss?

aiya who knows its all wierd >_> someone get a real japanese translator plx.

Malicx · 16 Aug 2006

Originally Posted by Arvec

Ah? They don't really make sense. Why would it be an addition of the mob evasion to the base hit chance? And why would shuriken mastery have a lower base chance to hit?

That's why it perplexes me. It also says something about a ratio (supposedly), which also doesn't make sense with the equation given. Anyway, I just thought it was interesting.

Kikkoman251 · 16 Aug 2006

Haha! Wooooohooooooooo.........I was right, all those Japanese/Korean guides that said AC affected SS accuracy was right. You guys could've just read foreign guides. =.=

P.S. I think that's the reason my Korean friend(who has played foreign Trickster before) recommended a 2341 build for my fox, that explains the 2.

oryx · 16 Aug 2006

so you're saying that lions should often or almost always hit with stone strike. can we have some confirmation on this lions?